Users typically visit enterprise interaction channels, such as an enterprise Website for example, to buy a good or a service, to learn about new good or service offerings, to troubleshoot a product, to lodge a complaint, and the like. Most enterprises, nowadays, track user actions on the enterprise interaction channels. For example, if a user is interacting with an interactive voice response (IVR) system for resolving a query, then the nature of the query, whether the query was resolved or not, and many such attributes may be tracked by the enterprise.
In a sales scenario, the tracked attributes corresponding to user actions on the enterprise interaction channels are typically provided as an input to a binary classifier to predict whether the user will purchase a good or a service or not. If the user is predicted to make a purchase transaction, then the user is classified as a hot lead, and if the user is predicted not to make a purchase transaction then the user is classified as a non-hot lead.
Typically, the users who are classified as hot leads are proactively targeted during their current journey on the enterprise interaction channels, whereas the non-hot leads are largely ignored or are targeted only if the resources are available after engaging with the hot leads. In an example scenario, a button chat may be displayed on an enterprise Website to a user categorized as a non-hot lead. The chat-based assistance may, in such a scenario, only be available to the user if the user seeks assistance by clicking or selecting the button chat option.
Generally, user categorization as a hot lead or a non-hot lead is performed using stringently applied boundaries and, as such, many users whose activity on the enterprise interaction channels may not be captured accurately or for whom sufficient data is not available for categorization purposes, are classified as a non-hot leads, and thereby ignored.
Accordingly, there is a need to drive more opportunity for sales by categorizing users in alternate ways. There is also a need to engage proactively with the non-hot leads because the non-hot leads constitute a substantially higher percentage of the overall pool of users, and although they are least likely to buy, they may still be proactively engaged on the enterprise interaction channels to drive up sales for an enterprise.